Prediction Market Liquidity Sourcing: Real Institutional Case Study
Institutional investors entering prediction markets face one core problem before anything else: liquidity. Without reliable liquidity sourcing, large positions create punishing slippage, price discovery breaks down, and the entire alpha thesis falls apart. This case study traces how one mid-sized systematic fund solved that problem — step by step — and what other institutional participants can learn from their approach.
Why Liquidity Is the Institutional Investor's First Problem in Prediction Markets
Prediction markets have matured dramatically since 2020. Platforms like Polymarket have settled billions in contract volume, Kalshi gained CFTC regulatory approval, and new entrants continue to emerge. Yet the market structure remains fundamentally thinner than traditional financial markets.
For retail traders moving $500 into a binary contract, this rarely matters. For an institutional desk moving $250,000 into the same contract, it matters enormously.
Bid-ask spreads on mid-tier prediction market events can range from 2% to 8% — compared to sub-0.1% on liquid equity markets. Order book depth is shallow: even on Polymarket's most-traded political events, the visible depth within 3% of mid-price often sits below $200,000. For an institution sizing positions at $500,000 or more, this creates a structural challenge that must be addressed at the portfolio construction stage, not as an afterthought.
The fund in this case study — a 12-person quantitative trading firm managing approximately $180M in alternative strategies — learned this lesson the expensive way before building a systematic liquidity-sourcing framework.
The Fund's Initial Foray: What Went Wrong First
In Q3 2024, the fund allocated $2M to prediction market strategies across three platforms. Their thesis was sound: they had built a proprietary model with documented edge on short-term political and macroeconomic binary outcomes.
The execution was disastrous.
On one flagship trade — a 65% probability contract they believed was mispriced at 58% — they attempted to deploy $400,000. Their average fill price came in at 63.2%, consuming most of their perceived edge before a single day had passed. Total slippage on their initial $2M deployment averaged 4.1%, compared to their modeled assumption of 0.8%.
Their backtested returns had assumed retail-level position sizing. Their actual position sizing was 20x to 40x larger than any single liquidity provider in their target markets was prepared to absorb.
This is a common failure mode. For a deeper look at how slippage interacts with prediction market position sizing, the advanced slippage strategies for prediction markets in Q2 2026 guide breaks down exactly how to model execution costs before entering a position.
Building a Liquidity Sourcing Framework: The Five-Step Process
After their initial losses, the fund spent three months building a structured liquidity sourcing process. Here is the exact methodology they developed:
Audit current market depth across target platforms. For every market the model flagged, traders were required to snapshot order book depth at 1%, 3%, and 5% from mid-price — logging it three times daily for two weeks before entry.
Classify contracts by liquidity tier. Contracts were bucketed into Tier 1 (>$500K visible depth within 5%), Tier 2 ($100K–$500K), and Tier 3 (<$100K). Position size limits were set at 15% of available depth per tier.
Identify and engage OTC liquidity providers. Rather than relying solely on on-chain order books, the fund built relationships with three dedicated prediction market market makers willing to provide indicative quotes for block trades outside the visible book.
Implement time-weighted entry algorithms. For Tier 1 markets, they broke entries into tranches over 6–72 hours depending on contract duration, targeting periods of historically high order flow (typically 9 AM–11 AM ET and 2 PM–4 PM ET).
Establish cross-platform arbitrage as a liquidity backstop. When on-chain liquidity was insufficient, the desk used price discrepancies across platforms to partially source fills — effectively treating arbitrage as a tool to aggregate liquidity rather than purely as a profit center.
OTC Liquidity: The Hidden Engine Institutions Actually Use
The single biggest discovery for the fund was that published order book depth is not the actual available liquidity. The visible book on any prediction market platform represents perhaps 30%–50% of total accessible liquidity for institutional participants.
The rest lives in:
- Direct market maker relationships — Several firms now specialize in prediction market market making. They maintain internal risk limits and will quote blocks ranging from $50,000 to $2,000,000 on major events, often at tighter spreads than the on-chain book.
- Prediction market desks at crypto-native trading firms — Firms like Wintermute and GSR have reported expanding into event derivatives. Their institutional desks can intermediate significant size on high-profile events.
- Cross-venue aggregation — Using platforms like PredictEngine that aggregate liquidity signals across venues allows institutional participants to identify where total available liquidity is deepest before committing to an entry strategy.
The fund ultimately sourced approximately 38% of their total prediction market volume through OTC channels by Q1 2025 — dramatically reducing average execution slippage from 4.1% to 1.2%.
Quantitative Results: Before and After Liquidity Framework
The following table summarizes the fund's performance metrics in their first six months of ad hoc trading versus their first six months after implementing the structured liquidity framework:
| Metric | Ad Hoc Period (Q3–Q4 2024) | Framework Period (Q1–Q2 2025) |
|---|---|---|
| Average slippage per trade | 4.1% | 1.2% |
| % of trades achieving modeled fill | 31% | 78% |
| Gross alpha (model edge) | 6.8% | 6.5% |
| Net alpha (after execution costs) | 2.7% | 5.3% |
| Largest single-trade slippage | 11.4% | 3.1% |
| OTC volume as % of total | 0% | 38% |
| Prediction market AUM allocated | $2.0M | $8.5M |
The numbers tell a clear story: execution quality improvement, not model improvement, drove a near-doubling of net returns. The model itself barely changed. Liquidity sourcing was the variable that transformed a marginal strategy into a high-conviction allocation.
Platform Selection and Regulatory Considerations for Institutions
Not all prediction market platforms are accessible to institutional participants, and those that are accessible carry different liquidity profiles and regulatory constraints.
Regulated vs. Unregulated Venues
Kalshi — the first CFTC-regulated prediction market exchange — offers institutional-grade infrastructure including limit orders, custodial integrations, and formal KYC/AML compliance. Liquidity on Kalshi tends to be thinner than Polymarket for most contracts, but the regulatory clarity is valuable for compliance-sensitive institutions. Understanding Kalshi limit orders and execution mechanics is essential before committing capital on that platform.
Polymarket — technically offshore and accessible to non-US participants — has substantially deeper liquidity on political and macroeconomic events. The fund in this case study used Polymarket for the majority of their volume but maintained strict internal compliance review on each market entered.
Emerging platforms — Several new entrants are targeting institutional liquidity specifically, some with RFQ (request-for-quote) systems modeled on traditional derivatives markets.
Key Compliance Steps for Institutions
Before deploying institutional capital, the fund required every team member to complete:
- Full KYC documentation on every platform used
- Legal opinion on each jurisdiction's treatment of prediction market contracts
- Risk committee sign-off on maximum allocation per platform
- Quarterly review of platform counterparty risk
Failing to address KYC and wallet infrastructure systematically is one of the most common and costly mistakes in this space — a topic covered in depth in this guide on KYC and wallet setup mistakes that cost prediction market traders.
Portfolio Integration: Hedging and Tax Efficiency at Scale
For institutional investors, prediction market positions don't exist in isolation. They need to fit within a broader portfolio framework, including hedging, correlation management, and tax reporting.
Hedging Prediction Market Exposure
The fund used prediction market positions as both alpha generators and portfolio hedges. A 70% probability contract on a rate-cut outcome, for example, was used to offset duration risk in their fixed income book. This cross-asset hedging function actually improved the case for prediction markets as an asset class — not just as a speculation vehicle.
For structured approaches to integrating prediction markets into a broader hedging strategy, the advanced portfolio hedging strategy and Q2 2026 predictions article provides a practical framework.
Tax Treatment of Prediction Market Profits
Tax reporting on prediction market income remains an area of significant complexity. Binary contracts settled in USDC or stablecoins, offshore platform usage, and the treatment of unrealized contract value all create reporting complications that traditional fund administrators are poorly equipped to handle.
The fund brought in a crypto-specialist accounting firm in Q4 2024, which proved essential. Resources like the crypto prediction markets tax guide with backtested results provide a useful starting framework for institutions beginning to size up these obligations.
What the Industry Is Getting Wrong: Institutional Liquidity Myths
Several persistent myths continue to distort how institutional investors approach prediction market liquidity sourcing.
Myth 1: "If I break my order into small pieces, slippage disappears." Partially true, but prediction markets have a different microstructure from equity markets. Small order splitting across thin books can actually signal directional intent, causing liquidity providers to widen quotes proactively.
Myth 2: "Polymarket has enough liquidity for any institutional trade." Polymarket routinely handles $1M+ single-event volume, but that volume is distributed across many participants over days or weeks — not available to a single institutional buyer entering at once.
Myth 3: "OTC prediction market liquidity doesn't exist." The fund's experience directly refutes this. OTC channels now represent a meaningful share of actual institutional volume in this market, and that share is growing rapidly as more sophisticated participants enter.
Myth 4: "Liquidity will improve on its own as markets grow." Liquidity in prediction markets is heavily event-specific. A U.S. presidential election contract and a county-level referendum contract occupy the same platform but operate in completely different liquidity environments. Institutions must evaluate each market independently — using tools like PredictEngine to assess real-time depth before sizing positions.
Frequently Asked Questions
What is prediction market liquidity sourcing?
Prediction market liquidity sourcing refers to the process of identifying, aggregating, and accessing available capital to buy or sell prediction market contracts at acceptable prices. For institutional investors, this includes both on-chain order books and off-exchange OTC channels with professional market makers.
How much liquidity is typically available on major prediction market platforms?
On Polymarket's most active markets — major U.S. elections or Federal Reserve decisions — total visible depth can reach $2M–$10M across the full order book. However, depth within 2–3% of the current mid-price is often significantly smaller, and institutions should expect meaningful slippage on trades above $100,000 without OTC sourcing.
Are prediction markets accessible to institutional investors in the United States?
Kalshi is CFTC-regulated and explicitly accessible to U.S. institutional participants. Polymarket restricts U.S. residents per its terms of service. Many institutions use legal and compliance counsel to navigate platform-specific rules before allocating capital, and some access offshore platforms through appropriately structured entities.
How do institutional investors reduce slippage in prediction markets?
The most effective techniques include time-weighted order splitting, engaging OTC market makers for block trades, cross-platform arbitrage to aggregate fills, and restricting position sizes to a defined percentage of available book depth. The fund in this case study reduced average slippage from 4.1% to 1.2% using these methods.
What role do market makers play in prediction market liquidity?
Market makers provide two-sided quotes — bids and offers — that give other participants the ability to enter and exit positions. In prediction markets, market makers often bear significant directional risk and price their quotes accordingly. For institutional block trades, dedicated market makers will often negotiate custom spreads outside the visible order book.
How should institutions account for prediction market profits and losses for tax purposes?
This depends heavily on jurisdiction, platform structure, and contract type. In the United States, gains from prediction contracts may be treated as ordinary income, Section 1256 contracts, or short-term capital gains depending on the platform and legal structure used. Institutions should engage tax counsel with specific crypto-derivatives experience before filing — not after.
Take the Next Step With PredictEngine
Sourcing liquidity in prediction markets at institutional scale is genuinely hard — but it is a solvable problem with the right data, relationships, and execution infrastructure. The fund profiled in this case study went from 2.7% net alpha to 5.3% net alpha without changing their model at all. The variable was execution quality, and execution quality starts with understanding where real liquidity lives.
PredictEngine is built specifically to help serious traders and institutions navigate prediction market complexity — from real-time liquidity analytics to cross-platform position monitoring and execution tools. Whether you are deploying $50,000 or $5,000,000, the right infrastructure makes the difference between capturing your edge and losing it to slippage before the market even moves. Visit PredictEngine today to explore how institutional-grade prediction market tools can transform your execution strategy.
Run your own quant desk
Felix lets you research, backtest and run automated trading strategies, with risk controls built in.
Try Felix